Optimization of 4D-CT in the Diagnostic Workup of Tumors of the Pancreas

Patients with painless jaundice require an ultrasound, CT and/or MRI of the
abdomen to diagnose the cause of obstruction. Possible causes are a stone or
tumor in the common bile duct, an inflammation or a cystic or solid tumor in
the head of the pancreas. In case of a tumor the exact location, its extent to
surrounding structures and the presence of locoregional and distant metastases
have to be determined.

Patients with a tumor in the pancreas are usually diagnosed with a CT of the
thorax and abdomen. The CT scan acquisition is composed of a non-contrast
enhanced series, followed by a contrast-enhanced series in the arterial phase
and in the porto-venous phase (Figure 1).

Recently a new CT scanner (Aquilion ONE, 320 row, Toshiba Medical Systems,
Ohtawara, Japan) has been installed at the Department of Radiology enabling
dynamic volumetric scan covering up to 16 cm in length, at minimum 0.35 seconds
rotation time; specifications that have been unmatched before. This new
technique permits dynamic volumetric imaging of the pancreas and liver, making
it possible to perform time-resolved functional studies (Figure 1). This may
improve early detection of liver metastases and improve detection and
characterization of the primary pancreatic tumor.

Until now, a diagnostic scan and perfusion CT scan are performed in two
separate steps, each requiring a contrast injection. With the new dynamic
volumetric scan technique both steps can be combined, resulting in one CT scan
with only one contrast injection and with both diagnostic and perfusion
characteristics available. However, as it is a new technique, it is necessary
to optimize the CT scan parameters, regarding respiration technique, the number
of necessary time points in the arterial and porto-venous phase and radiation
dose per time point.

When compared to a normal diagnostic scan, the data obtained with a perfusion
scan are likely to improve the detection and characterization of the pancreatic
tumor. This will lead to a better staging of the tumor resulting in a better
selection of therapy, either curative resection (Whipple operation) or
palliative chemo-radiation therapy. With better staging non-curative resections
can be avoided, reducing morbidity and increasing quality of life.

--------------------------------------------------------------------------------
------------------------------
Fig. 1. Time density curve. Contrast enhancement in Hounsfield units (HU) as a
function of time (t) in seconds (s). A CT scan-protocol in which a diagnostic
and perfusion scan are performed during one intravenous contrast injection.
Phase I, II and III are the non-contrast enhanced (t=0s), and arterial (t=35s)
and porto-venous (t=65s) contrast-enhanced diagnostic scans respectively. The
curve (red) represents the data acquired at regular time intervals during the
perfusion scan.

With this pilot study our aim is to lower the radiation dose. Therefore we will
scan 5 patients at 100kV and 5 patients at 80kV and determine the resultant
decrease in radiation dose, aiming at a dose comparable with a normal
diagnostic CT scan with 3 phases. This can be achieved by optimizing the number
of time points, the dose per time point and combining perfusion data to create
diagnostic data. In the latter case, less diagnostic scans are necessary. The
use of adaptive iterative dose reduction (AIDR 3D) will reduce the radiation
dose even more. This technique just recently became available (in October 2011).

The acquired data will result in an optimized CT scan protocol regarding the
minimally required data to perform a time-based functional study with the
unique capability to combine it with a diagnostic scan and with it a as low as
possible radiation dose. The final aim is to improve staging and
characterization of pancreatic tumors resulting in less non-curative resections
with associated postoperative morbidity and improve quality of life.

Number of included patients: n=10.

Inclusion criteria:
- Patients with a suspected tumor in the head of the pancreas.
- Older than 18 years

Exclusion criteria:
- Non post menopausal women
- Transpapillary endoprosthesis in situ
- Previous surgery of the pancreas
- Previous chemoradiation therapy of the pancreas
- Patients with active hepatitis or other hepatic diseases that may cause
jaundice
- Informed consent not obtained

Respiration protocol:
A band will be applied around the upper abdomen to restrict movements of the
abdominal wall during normal respiration. Before the start of the scan the
patient is asked to hyperventilate for a minute, in order to obtain a quiet and
even respiration during scanning.

Specifications of contrast administration:
- Contrast agent: Xenetix 300mgI/ml
- Amount contrast agent: 0,5 - 0,6g I/kg bodyweight
- Injection rate: 3-10ml/s
- No delay between start of scan and start of injection
With the new CT scan the amount of contrast agent, used for combined perfusion
and diagnostic imaging, is the same as that of a standard diagnostic scan. The
renal function of patients will be monitored according to the current standards
for the administration of intravenous contrast agents.

Specifications of radiation dose (estimated):
- Normally: a diagnostic CT scan with 3 phases (kV = 120; mA variable): 12-26
mSv
o at T = 0s: non-contrast enhanced helical CT upper abdomen
o at T = 35s: contrast-enhanced helical CT upper abdomen
o at T = 65s: contrast-enhanced helical CT thorax and abdomen
- Study: a diagnostic CT scan with 3 phases and perfusion CT scan (kV = 80/100;
mA variable): 36-50 mSv
o at T = 0s: non-contrast enhanced helical CT upper abdomen
o T = 0 - 300s (perfusion): contrast-enhanced upper abdomen
o at T = 35s: contrast-enhanced helical CT upper abdomen
o at T = 65s: contrast-enhanced helical CT thorax and abdomen
The extra radiation dose with the scan technique in this pilot is high, and
doubles the effective dose for the patient (European Guidelines on Quality
Criteria for Computed Tomography, EUR 16262 EN, Table 2, p70). The perfusion
scan of the upper abdomen will result in an additional radiation dose of 24
mSv. The additional risk for this specific oncologic patient group is almost
negligible. Patients diagnosed with pancreatic cancer have a 1 year survival of
20% and a 5 year survival of 5%. For local disease the 5 year survival is
approximately 20%, while the median survival for locally advanced and for
metastatic disease is about 10 and 6 months respectively. Five year survival
after curative resection of a adenocarcinoma in the head of the pancreas is
5-10%. Radiation-induced cancer has a latency period that substantially
exceeded 5 years.

With the already implemented new software (AIDR-3D) the total estimated
radiation dose is reduced to 20-36mSv (see paragraph on risks associated with
participation).

Data analysis:
- The results of the diagnostic CT scans are part of the regular workup for
patients with a pancreatic tumor and will be discussed in the weekly pancreatic
tumor working group (PACON), resulting in a treatment plan.
-The perfusion data will be analyzed separately.
oSoftware for registration to compensate for possible movement during scanning
and perfusion analysis, as available on the Toshiba workstation, will be
compared with software analysis tools developed in-house by DIAG using MeVisLab
software development environment.
oThe number of time points will be optimized regarding radiation dose and
perfusion parameters.
oTo combine perfusion data to create a diagnostic scan. Omission of a
diagnostic scan will result in a reduction of the radiation dose.
oWith the use of iterative reconstruction, a new technique to reconstruct
images, image quality (signal to noise ratio (SNR)), can be improved. This data
are used to calculate dose reduction protocols for future patients.

The risk associated with participation in this study is the extra radiation
dose involved in the additional perfusion scan.

With the recent implementation of adaptive iterative dose reduction (AIDR-3D)
software in October 2011, the radiation dose can be diminished. We tested the
new software on a phantom (CT Torst Phantom CTU-41, Kyoto Kagaku Co., Ltd) with
the CT scan protocol as described above. The resulting estimated dose reduction
depending on the size of the patient (20
Old software (QDS+):
- diagnostic scan : 12-26mSv (20 - perfusion scan (100kV) : 24mSv
TOTAL: 36-50mSv

New software (AIDR-3D):
- diagnostic scan : 4-20mSv (20 - perfusion scan (100kV): 16mSv
TOTAL: 20-36 mSv